2025-08-26 マックス・プランク研究所
<関連情報>
- https://www.mpg.de/25232729/0821-defo-cohort-life-expectancy-is-no-longer-rising-as-quickly-154642-x
- https://www.pnas.org/doi/10.1073/pnas.2519179122
集団の平均寿命はもはや急速に伸びていない Cohort mortality forecasts indicate signs of deceleration in life expectancy gains
José Andrade, Carlo Giovanni Camarda, and Héctor Pifarré i Arolas
Proceedings of the National Academy of Sciences Published:August 25, 2025
DOI:https://doi.org/10.1073/pnas.2519179122
Significance
Longevity, measured by cohort life expectancy, has risen steadily in high-income countries since the early 1900s, raising expectations that this trend will continue. However, whether these gains will persist remains uncertain. Using multiple mortality forecasting methods, we find robust evidence that currently living cohorts, born between 1939 and 2000, may experience substantially slower gains. This slowdown is driven mainly by a diminished pace of mortality improvements at very young ages, with over half of the deceleration linked to trends under age 5. As this pattern is already evident in observed data, even if our forecasts are overly pessimistic, a reversal is unlikely. These findings have important implications for future longevity projections and social policy.
Abstract
The fast-paced improvements in mortality in high-income countries since the early 1900s have led to a sustained increase in life expectancy. However, whether this linear trend will continue or life expectancy gains will decelerate in the near future remains unclear. To answer this question, we apply multiple established and recently developed mortality forecasting methods to estimate cohort life expectancy for individuals born between 1939 and 2000 in 23 high-income countries. Across all forecasting methods, our results robustly and consistently indicate a deceleration in cohort life expectancy. The previously observed pace of improvement, 0.46 y per cohort, declines by 37% to 52%, depending on the method used. Robustness checks suggest that these findings are unlikely to be solely due to downward bias in cohort life expectancy forecasts. Furthermore, an age-decomposition analysis indicates that this deceleration is primarily driven by a slower pace of mortality improvement at very young ages. Over half of the total deceleration is attributable to mortality trends under age 5, while more than two-thirds is explained by mortality trends under age 20. This pattern had already emerged in the observed data for the cohorts included in our analysis. Thus, even if these estimates turned out to be overly pessimistic, it is unlikely that the deceleration will reverse in the near future.
